interwoven three-dimensional architecture of cobalt oxide...

8
Interwoven Three-Dimensional Architecture of Cobalt Oxide Nanobrush-Graphene@Ni x Co 2x (OH) 6x for High-Performance Supercapacitors Longbing Qu, Yunlong Zhao, ,§ Aamir Minhas Khan, Chunhua Han,* ,Kalele Mulonda Hercule, Mengyu Yan, Xingyu Liu, Wei Chen, Dandan Wang, Zhengyang Cai, Wangwang Xu, Kangning Zhao, Xiaolin Zheng, and Liqiang Mai* ,State Key Laboratory of Advanced Technology for Materials Synthesis and Processing, Wuhan University of Technology, Wuhan 430070, China Department of Mechanical Engineering, Stanford University, Stanford, California 94305, United States § Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, United States * S Supporting Information ABSTRACT: Development of pseudocapacitor electrode materials with high comprehensive electrochemical performance, such as high capacitance, superior reversibility, excellent stability, and good rate capability at the high mass loading level, still is a tremendous challenge. To our knowledge, few works could successfully achieve the above comprehensive electrochemical performance simultaneously. Here we design and synthesize one interwoven three-dimensional (3D) architecture of cobalt oxide nanobrush-graphene@Ni x Co 2x (OH) 6x (CNG@ NCH) electrode with high comprehensive electrochemical performance: high specic capacitance (2550 F g 1 and 5.1 F cm 2 ), good rate capability (82.98% capacitance retention at 20 A g 1 vs 1 A g 1 ), superior reversibility, and cycling stability (92.70% capacitance retention after 5000 cycles at 20 A g 1 ), which successfully overcomes the tremendous challenge for pseudocapacitor electrode materials. The asymmetric supercapacitor of CNG@NCH//reduced-graphene- oxide-lm exhibits good rate capability (74.85% capacitance retention at 10 A g 1 vs 0.5 A g 1 ) and high energy density (78.75 Wh kg 1 at a power density of 473 W kg 1 ). The design of this interwoven 3D frame architecture can oer a new and appropriate idea for obtaining high comprehensive performance electrode materials in the energy storage eld. KEYWORDS: Three-dimensional frame architecture, cobalt oxide nanobrush-graphene@nickle-cobalt hydroxides, comprehensive performance, tremendous challenge, pseudocapacitor materials A t the forefront of electrical energy storage systems, especially for hybrid electric vehicles, smart grids, and the portable electronics, there are batteries and electrochemical capacitors (ECs). 15 Compared with batteries, ECs have attracted great interest due to their desirable properties, such as fast charge and discharge, good cycle performance, and high power density. 1,6,7 Generally, there exist two types of supercapacitors based on the charge storage mechanism. The rst type is electrochemical double-layer capacitors (EDLCs), which store the charge electrostatically by reversible adsorption of ions in the electrolyte on the surface of active materials. 6,810 The second type of ECs, known as the pseudocapacitors (PCs), use fast and reversible surface or near-surface reactions for charge storage. 6,8 According to the two dierent storage mechanisms, the PCs display higher capacitance and energy density, but poorer rate capability and cycle stability than the EDLCs. 4,5 Among the PC materials, Co 3 O 4 has been regarded as a scalable alternative material due to its high conductivity and high theoretical specic capacitance. 5,1113 Graphene with extraordinarily high electrical conductivity is an excellent material to increase the conductivity of PC materials. 14,15 In addition, transition metal hydroxides, especially Ni(OH) 2 and Co(OH) 2 , have been widely studied owing to their abundance, easy preparation, and well-dened electrochemical redox activity. 1620 Ni(OH) 2 has a high theoretical speci c capacitance, but its poor intrinsic conductivity restricts its rate capability. 19,21 On the other hand, Co(OH) 2 possesses higher conductivity than Ni(OH) 2 . 22,23 Therefore, the integration of Ni and Co hydroxides exhibit enhanced specic capacitance and rate capability. 17,21,24,25 To synergistically improve the ion diusion kinetics and electron transport, the eective strategy tends to combine dierent electrode materials in various Received: December 21, 2014 Revised: February 13, 2015 Published: February 24, 2015 Letter pubs.acs.org/NanoLett © 2015 American Chemical Society 2037 DOI: 10.1021/nl504901p Nano Lett. 2015, 15, 20372044

Upload: others

Post on 07-Jul-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Interwoven Three-Dimensional Architecture of Cobalt OxideNanobrush-Graphene@NixCo2x(OH)6x for High-PerformanceSupercapacitorsLongbing Qu,† Yunlong Zhao,†,§ Aamir Minhas Khan,† Chunhua Han,*,† Kalele Mulonda Hercule,†

Mengyu Yan,† Xingyu Liu,† Wei Chen,† Dandan Wang,† Zhengyang Cai,† Wangwang Xu,†

Kangning Zhao,† Xiaolin Zheng,‡ and Liqiang Mai*,†

†State Key Laboratory of Advanced Technology for Materials Synthesis and Processing, Wuhan University of Technology, Wuhan430070, China‡Department of Mechanical Engineering, Stanford University, Stanford, California 94305, United States§Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, United States

*S Supporting Information

ABSTRACT: Development of pseudocapacitor electrode materials with highcomprehensive electrochemical performance, such as high capacitance, superiorreversibility, excellent stability, and good rate capability at the high mass loadinglevel, still is a tremendous challenge. To our knowledge, few works couldsuccessfully achieve the above comprehensive electrochemical performancesimultaneously. Here we design and synthesize one interwoven three-dimensional(3D) architecture of cobalt oxide nanobrush-graphene@NixCo2x(OH)6x (CNG@NCH) electrode with high comprehensive electrochemical performance: highspecific capacitance (2550 F g−1 and 5.1 F cm−2), good rate capability (82.98%capacitance retention at 20 A g−1 vs 1 A g−1), superior reversibility, and cyclingstability (92.70% capacitance retention after 5000 cycles at 20 A g−1), whichsuccessfully overcomes the tremendous challenge for pseudocapacitor electrodematerials. The asymmetric supercapacitor of CNG@NCH//reduced-graphene-oxide-film exhibits good rate capability (74.85% capacitance retention at 10 A g−1

vs 0.5 A g−1) and high energy density (78.75 Wh kg−1 at a power density of 473 W kg−1). The design of this interwoven 3Dframe architecture can offer a new and appropriate idea for obtaining high comprehensive performance electrode materials in theenergy storage field.

KEYWORDS: Three-dimensional frame architecture, cobalt oxide nanobrush-graphene@nickle-cobalt hydroxides,comprehensive performance, tremendous challenge, pseudocapacitor materials

At the forefront of electrical energy storage systems,especially for hybrid electric vehicles, smart grids, and

the portable electronics, there are batteries and electrochemicalcapacitors (ECs).1−5 Compared with batteries, ECs haveattracted great interest due to their desirable properties, suchas fast charge and discharge, good cycle performance, and highpower density.1,6,7 Generally, there exist two types ofsupercapacitors based on the charge storage mechanism. Thefirst type is electrochemical double-layer capacitors (EDLCs),which store the charge electrostatically by reversible adsorptionof ions in the electrolyte on the surface of active materials.6,8−10

The second type of ECs, known as the pseudocapacitors (PCs),use fast and reversible surface or near-surface reactions forcharge storage.6,8 According to the two different storagemechanisms, the PCs display higher capacitance and energydensity, but poorer rate capability and cycle stability than theEDLCs.4,5

Among the PC materials, Co3O4 has been regarded as ascalable alternative material due to its high conductivity and

high theoretical specific capacitance.5,11−13 Graphene withextraordinarily high electrical conductivity is an excellentmaterial to increase the conductivity of PC materials.14,15 Inaddition, transition metal hydroxides, especially Ni(OH)2 andCo(OH)2, have been widely studied owing to their abundance,easy preparation, and well-defined electrochemical redoxactivity.16−20 Ni(OH)2 has a high theoretical specificcapacitance, but its poor intrinsic conductivity restricts its ratecapability.19,21 On the other hand, Co(OH)2 possesses higherconductivity than Ni(OH)2.

22,23 Therefore, the integration ofNi and Co hydroxides exhibit enhanced specific capacitanceand rate capability.17,21,24,25 To synergistically improve the iondiffusion kinetics and electron transport, the effective strategytends to combine different electrode materials in various

Received: December 21, 2014Revised: February 13, 2015Published: February 24, 2015

Letter

pubs.acs.org/NanoLett

© 2015 American Chemical Society 2037 DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

architecture designs.26,27 In the energy storage field, manyresearchers have focused on transition metal oxides (MnO2,

28

RuO2,29 Co3O4,

30 NiO,31 and so forth) combined with othermetal oxides, metal hydroxides, carbonaceous and conductingpolymers in various architecture designs to obtain thepseudocapacitor electrode materials with comprehensiveelectrochemical performance and achieved significant re-sults.32−37 Especially, Zheng’s group created CoNiO2/TiN−TiOxNy nanostructured composite electrode that exhibitedultrahigh gravimetric specific capacitance (3181 F g−1 at currentdensity of 2 mA cm−2).38 However, development ofpseudocapacitor electrode materials with high comprehensiveelectrochemical performance, such as high capacitance, superiorreversibility, excellent stability, and good rate capability at thehigh mass loading level, still is a tremendous challenge, andmost of these works hardly obtain the pseudocapacitormaterials with all the above high comprehensive electro-chemical performance simultaneously.In the present work, we designed and elaborated one special

interwoven three-dimensional (3D) frame architecture ofcobalt oxide nanobrush-graphene@NixCo2x(OH)6x (CNG@NCH) electrode through a facile modified hydrothermalmethod followed by two-step calcination and two-stepoptimized electrodeposition39−41 (Figure 1b). The distinctionof our design strategy with the previous strategies is that weelaborately design interwoven 3D frame architecture with verythin graphene film decorated on the surface of cobalt oxidenanobrush, which is composed by numbers of ultrathinnanosheets, nickel−cobalt hydroxide nanoflakes uniformlyimmobilized and firmly interconnected with cobalt oxidenanobrush-graphene, which enables this 3D frame architectureelectrode to inherit the high electrochemical activity from thefull utilization of both Co3O4 nanobrush (NB) andNixCo2x(OH)6x (NCH) nanoflake, the high electronicconductivity from graphene. Meanwhile, this 3D framearchitecture neither prevents OH− from contacting directlywith CNG nor increases the ion diffusion path. All of thesemerits of this interwoven 3D frame architecture make our

present electrode have self-adaptive strain-relaxation capabilityduring the charge and discharge processes, fast electrontransport, and ion diffusion, simultaneously (Figure 1a) thusleading to high comprehensive performance of the present 3Dframe architecture electrode: high specific capacitance (2550 Fg−1 and 5.1 F cm−2), good rate capability (82.98% capacitanceretention at 20 A g−1 vs 1 A g−1), superior reversibility, andcycling stability (92.70% capacitance retention after 5000 cyclesat 20 A g−1). More importantly, our present interwoven 3Dframe architecture CNG@NCH successfully overcomes thetremendous challenge for pseudocapacitor electrode materialsand this high level comprehensive performance has not beensuccessfully achieved before. To further manifest the potentialapplication of this 3D frame architecture electrode, wefabricated an asymmetric supercapacitor by combining the 3Dframe architecture electrode with the reduced-graphene-oxide-film (RGOF) electrode. The asymmetric supercapacitorexhibits good rate capability (74.85% capacitance retention at10 A g−1 vs 0.5 A g−1), high energy density (78.75 Wh kg−1 at apower density of 473 W kg−1), and excellent cycle capability(99.15% capacitance retention after 5000 cycles at 10 A g−1).Therefore, our 3D frame architecture material successfullyovercomes the tremendous challenge for pseudocapacitorelectrode materials with high comprehensive electrochemicalperformance. It also can provide a promising structure designdirection for optimizing the electrochemical performance ofvarious electrode materials and could be generally applicable tothe future practical application of electrochemical energystorage.The schematic synthesis procedure of our interwoven 3D

frame architecture of CNG@NCH electrode is illustrated inFigure 1b. The X-ray diffraction (XRD) patterns of the sampleswere collected by using a Bruker D8 Advance X-raydiffractometer with a nonmonochromated Co Ka X-ray source.The peaks of Co3O4 nanobrush (CN) are consistent very wellwith cubic Co3O4 (JCPDS card No. 80-1542) with an Fd3 ̅mspace group and lattice parameters of a = b = c = 8.1099 Å(Figure 1c). The Fourier transform infrared spectroscopic

Figure 1. (a) Charge storage mechanism. (b) Schematic illustration of the synthesis procedure of interwoven 3D frame architecture of CNG@NCH.(c) XRD patterns of NCH, CN and CNG@NCH, respectively. (d) FTIR spectra of CN and CNG. (e1) EDS mappings of CNG. (e2) EDS linescans of CNG showing the scanning route along the perpendicular direction (light blue line), the element distribution of carbon (yellow line), cobalt(deep blue line), and oxygen (red line), respectively. (e3) EDS line scans of CNG, showing the scanning route along the axial direction (light blueline), the element distribution of carbon (yellow line), cobalt (deep blue line), and oxygen (red line), respectively.

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2038

(FTIR) spectra (Figure 1d) provide evidence for the presenceof graphene. The bands at of CNG 1060, 1385, 1430, and 1620cm−1 are ascribed to the stretching vibration of C−O, C−OH,C−H, and CC, respectively.42,43 The energy dispersivespectroscopy (EDS) elemental mappings and line scans (Figure1e) clearly show the existence and distribution of graphene filmin Co3O4 nanobrush-graphene (CNG). The diffraction peaks ofthe XRD patterns for the as-prepared Ni−Co hydroxides(Figure 1c) match with those of α-Ni(OH)2 (JCPDS card No.38-0715) and hexagonal Co(OH)2 (JCPDS card No. 30-0443).The Ni−Co hydroxides are converted into NiCo2O4

(Supporting Information Figure S1b) after calcination,

indicating that the molar ratio of Ni and Co in the hydroxidesis 1:2, which is further verified by the atomic absorptionspectrometry (AAS) results (Supporting Information FigureS1c) of Ni−Co hydroxides. The diffraction peaks of the 3Dframe architecture material (Figure 1c) are consistent withthose of both CN and NCH and the existence of graphene isconfirmed by the FTIR and EDS measurements. All of theseresults indicate that the final sample is [email protected] electron microscopic (SEM) and transmission

electron microscopic (TEM) images are provided to show themorphology and structure of the as-prepared samples. From thehigh-magnification SEM image (Figure 2a), we can see that the

Figure 2. (a−e) SEM images (the inset shows the low magnification) of CN (a), CNG (b), NCH (c), and CNG@NCH (d,e) (high-magnificationand low-magnification of CNG@NCH), respectively. (f−h) TEM images of CNG@NCH (f), CNG (g), and CN (h), respectively. The inset of hand f show the HRTEM of CN and CNG@NCH, respectively.

Figure 3. (a) CV curves of the 3D frame architecture electrode at different scan rates. (b) Charge and discharge curves of the 3D frame architectureelectrode (CNG@NCH) at different current densities. (c) Specific capacitances of CN, CNG, NCH, CN@NCH, and CNG@NCH at differentcurrent densities, respectively. (d) Cycling performance of CN, CNG, NCH, and CNG@NCH at 20 A g−1 for 5000 cycles, respectively.

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2039

CN is composed by numbers of ultrathin nanosheets (about 15nm in thickness). From the high-magnification SEM and TEMimages (Figure 2b,g and Supporting Information Figure S3), weclearly observe that the graphene film with a thickness ofseveral nanometers is decorated on the surface of CN, which isin agreement with the EDS elemental mappings and line scansof CNG. The electrodeposition of graphene film on the surfaceof metal-oxide has been reported before.44 The high-magnification SEM and TEM investigations (Figure 2d,f) ofthe 3D frame architecture material reveal that the NCHnanoflakes are electrodeposited and well interweaved withCNG. The growth mechanism of the NCH nanoflakesinterweaved with CNG can probably be explained by“orientated attachment” and “self-assembly” as demonstratedin our previous work.37,45 In the high-resolution TEM image ofNCH (Figure 2f), the lattice fringe spacings of 2.32 and 2.37 Åcorrespond to the interplane distance of (015) and (101) of α-Ni(OH)2 and hexagonal Co(OH)2, respectively.To explore the electrochemical performance of the as-

prepared materials, we carried out the electrochemical testsusing a three-electrode configuration with a Pt plate counterelectrode and an Ag/AgCl reference electrode in 1 M LiOHaqueous electrolyte.46 Figure 3a presents the cyclic voltamme-try (CV) curves of the 3D frame architecture electrode atvarious scan rates; all CV curves display a similar shape that aresymmetric, indicating that this 3D frame architecture electrodehas a good electrochemical capacitive characteristic andsuperior reversible redox reaction. When the scan rates are 1,2, and 5 mV s−1, two pairs of reversible redox peaks areobserved. However, when the scan rate increases to 10 and 20mV s−1, only one pair of well-defined redox peaks are observed.This is because there is the superposition of the redox peaksand the polarization of this 3D frame architecture electrodewith the increase of scan rates.24,25 Remarkably, the peakpotential shifts only 100 mV for a 20 times increase of the scanrate, indicating that this 3D frame architecture electrodepossesses low polarization, which is in line with our expectation.As the graphene functions as an electron transport highway forboth CN and NCH nanoflakes, thus the 3D frame architectureelectrode presents an excellent electronic conductivity that isfurther verified by the electrochemical impedance spectra (EIS)results (Supporting Information Figure S6b). Additionally, theelectrochemical stability window of water and the potential forthe emergence of oxygen under our experimental conditionsare 0.52 V versus Ag/AgCl at the scan rate of 1 mV s−1

(Supporting Information Figure S5c). The nickel foam (1 cm ×1 cm) was also tested through CV at 5 mV s−1. Evidently,compared to that of CNG@NCH at the same scan rate, the CVintegrated area of the nickel foam is negligible (SupportingInformation Figure S5b).The charge−discharge (C-DC) testing (Figure 3b) shows

similar C-DC curves from −0.15 to 0.45 V over a wide range ofcurrent densities. The specific capacitances were calculatedbased on the total mass of the electrode materials loaded onnickel foam (0.75 mg cm−2 for CNG, 1.25 mg cm−2 for NCHnanoflake). According to the equation, C = 2(∫ I·V(t)dt)/(m·ΔV2) (eq 1),47 where I is the discharge current, V(t) isdischarge voltage excluding the IR drop, dt is the timedifferential, m is the mass loading of the electrode, and ΔV isthe voltage window, by replacing m by s, we can obtain the arealcapacitance of the electrode; the 3D frame architectureelectrode exhibits gravimetric specific capacitances of 2550,2480, 2350, 2230, and 2116 F g−1 at current densities of 1, 2, 5,

10, and 20 A g−1, respectively. The corresponding arealcapacitances are 5.1, 4.96, 4.7, 4.46, and 4.23 F cm−2 withCoulombic efficiency of 98.7, 99.2, 99.6, 100, and 100%,respectively. Over this current density range, the specificcapacitance maintains 82.98% of its initial value.Figure 3c displays the rate capabilities of CN, CNG, NCH,

CN@NCH, and CNG@NCH. Compared to CN, which has84.02% capacitance retention, the rate capability of CNG doesnot increase greatly although its gravimetric specific capacitancedecreases from 2579 to 2355 F g−1 corresponding to 91.31%capacitance retention when the current densities vary from 1 to20 A g−1. This is mainly because on the one hand the very thingraphene film decorated on the surface of CN increases theelectronic conductivity of CN. On the other hand, it preventsOH− from contacting directly with CN and increases the iondiffusion path. This is proved by the EIS results (SupportingInformation Figure S6b and S6b1). The CNG@NCH exhibitsa gravimetric specific capacitance decrease from 2550 F g−1 at 1A g−1 to 2116 F g−1 at 20 A g−1 with 82.98% capacitanceretention ,which is smaller than that of CNG. This can beexplained by the fact that the conductivity of NCH nanoflakesis lower than that of CNG,32 electrodeposition of NCHnanoflakes interweaved with CNG; in fact, it increases the Rs(which includes the inherent resistance of the electroactivematerial, ionic resistance of electrolyte, and contact resistance atthe interface between electrolyte and electrode) and Rct (chargetransfer resistance) that is demonstrated by the EIS results(Supporting Information Figure S6b), thus, leading to thedecrease of rate capability. It is worthy to note that although themass loading of the 3D frame architecture electrode (2.0 mgcm−2) is much larger than that of CNG (0.75 mg cm−2) this 3Dframe architecture electrode still has a specific capacitance of2550 F g−1, which is very close to that of CNG (2579 F g−1) at1 A g−1. This is mainly due to the special design of this 3Dbicontinuous frame architecture that the NCH nanoflakesuniformly immobilize and firmly interconnect with CNG,enabling both CNG and NCH nanoflakes to directly contactwith OH− in the electrolyte, largely opening the nanosizedactive sites to OH−. Compared to the hybrid electrode with theNCH coated on the surface of CNG (Supporting InformationFigure S8) and CN@NCH electrode, CNG@NCH electrodehas a much higher specific capacitance and rate capability (atthe same mass loading level). The gravimetric specificcapacitance of the hybrid electrode with the NCH coated onthe surface of CNG and CN@NCH electrode decrease from1553 to 1126 and 1947 to 1201 F g−1 when the currentdensities vary from 1 to 20 A g−1, which corresponds to 72.5and 61.7% capacitance retention, respectively. This is becausethe NCH nanoflakes firmly interconnect with CNG and are notjust coated on the surface of CNG, which leads to this 3Dframe architecture neither prevent OH− from contactingdirectly with CNG nor increase the ion diffusion path andthe graphene film as the middle part of the electrode provides ahighway of electron transport for CN and NCH (SupportingInformation Figure S6). All of these merits lead to a muchhigher capacitance and rate capability of this special 3D framearchitecture electrode. It should be noted that the arealcapacitance is also an important factor to evaluate theapplication of supercapacitors. From Supporting InformationFigure S7b, the areal capacitance of CNG@NCH is 5.1, 4.96,4.7, 4.46, and 4.22 F cm−2 at current densities of 1, 2, 5, 10, and20 A g−1, respectively. Moreover, we also summarized theelectrochemical performance of each component, CN@NCH

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2040

and CNG@NCH in Supporting Information Table S1, whichshows that the specific capacitance of our 3D frame architectureelectrode is higher than each component (CN, graphene, andNCH) and the areal capacitance of our 3D frame architectureelectrode is higher than the total areal capacitance of CN,graphene, and NCH at the same current density. Meanwhile,according to the equation of E = (∫ I·V(t)dt)/(m) and P= E/Δt(eq 2),48 where E is the energy density, P is the power density, Iis the discharge current, V(t) is the discharge voltage excludingthe IR drop, dt is the time differential, and Δt is the dischargetime, we calculated the energy density of this 3D framearchitecture electrode at the current density of 1 A g−1 with thepotential window of 1.5 V (Supporting Information Figure S9).This 3D frame architecture electrode exhibits an energy densityof 534.33 Wh kg−1, which is very close to the value ofcommercial lithium ion battery material LiFePO4 (maximaltheoretical energy density is 578 Wh kg−1).Figure 3d displays the cycling performance of CN, CNG,

NCH, and CNG@NCH at a current density of 20 A g−1. After5000 cycles, this 3D frame architecture electrode (CNG@NCH) possesses 92.70% capacitance retention, which is muchhigher than those of CN (52.60%), CNG (63.9%), and NCH(69.10%). It clearly exhibits that Ni−Co hydroxide nanoflakesfirmly interconnect with CNG materials, which can heavilyimprove the overall cycling stability. This is because nickel−cobalt hydroxide nanoflakes are uniformly immobilized andfirmly interconnected with the cobalt oxide-graphene ultrathinnanosheets to form the 3D frame architecture that has the self-adaptive strain-relaxation capability during the charge anddischarge processes, thus leading to the free-standing structurewithout self-aggregation. The SEM images after 5000 cycles(Supporting Information Figure S10) were captured for all theelectrode materials. It can be clearly observed that the 3D framearchitecture maintains a more stable morphology than those ofCN, CNG, and NCH nanoflakes after 5000 cycles, whichindicates that the structural destruction caused by the reactionwith the electrolyte is mitigated in the novel 3D framearchitecture electrode. The EIS results (Supporting InformationFigure S12) clearly show that the Rct of the 3D framearchitecture increases from 0.047 to 0.12 Ω, (Rs, 0.63 Ω vs 0.64Ω) after 5000 cycles. It indicates that the capacitance loss ismainly because of the morphology destruction and Rct increaseof this 3D frame architecture electrode.According to the above results, this 3D frame architecture

electrode has high capacitance, superior reversibility, good ratecapability, high energy density, and long cycle life, simulta-neously, which is mainly due to the special design of this 3Dframe architecture. This unique design has apparent merits asfollows (Figure 1a): (1) Compared with common slurry-coating technology, where a large surface area of active materialis blocked by the binders and the destruction of the structure ofactive materials during the slurry-coating process, the currentdesign is binder free and without structure damage of the activematerials. (2) The active material directly grows on the currentcollector, which provides fast electron transport and favorableelectrolyte accessible to the active materials. Moreover, it ismuch more direct and correct to study the relationship betweenthe structure and the performance of active material. (3) Themorphology of as-synthesized cobalt oxide is nanobrush, whichis composed by numbers of ultrathin nanosheets (about 15 nmin thickness), it has a shorter path for ion diffusion, and largelyavoids the “dead mass” of the active materials.20,49 From theSupporting Information Figure S4, the Co3O4 NB has a much

higher gravimetric specific capacitance and better rate capabilitythan those of Co3O4 nanowire and Co3O4 nanotube, whichfurther confirms the merits of the nanobrush morphology. (4)The very thin graphene film, which is the middle part in thisinterwoven 3D frame architecture, can sharply increase theconductivity of CN and it works as the electron transporthighway for both the CN and NCH nanoflakes. (5) Thestructure is 3D bicontinuous frame architecture and NCHnanoflakes firmly interconnect with the CNG materials withlarge surface areas accessible to electrolyte, enabling fullutilization of NCH, fast electron transport, and ion diffusionin this 3D structure, simultaneously. Moreover, this 3D framearchitecture neither prevents OH− from contacting directlywith CNG nor increases the ion diffusion path. The muchhigher capacitance and rate capability of CNG@NCH(compared to the hybrid electrode with the NCH coated onthe surface of CNG and CN@NCH electrode) further confirmit. (6) The 3D frame architecture electrode mitigates thestructural destruction during charge and discharge processes.The much good stability of CNG@NCH (compared to CN,CNG, and NCH) demonstrates it. Therefore, our 3D framearchitecture electrode based on the present design exhibits highcomprehensive electrochemical performance, which is betterthan other nanostructure electrodes that have been reportedbefore (Supporting Information Table S2).In order to further evaluate the potential application of this

3D frame architecture electrode, a two-electrode system wasused to measure the electrochemical performance of theCNG@NCH based asymmetric supercapacitor. Here the 3Dframe architecture electrode works as the positive electrode andthe RGOF works as the negative electrode, which the RGOFwas fabricated by using the method that has been reportedbefore.50,51 The electrolyte is 1 M LiOH aqueous solution. Allthe operating current densities are based on the total mass ofactive materials (CNG@NCH and RGOF). In this two-electrode system, the mass ratio of the positive to negativeelectrode is obtained by using the following equation: m+/m− =(C−·ΔV−)/(C+·ΔV+) (eq 3),52 (where m+ and m− are the massloading of the positive and negative electrodes, respectively, C+and C− are the gravimetric capacitance of the positive andnegative electrodes, respectively. ΔV+ and ΔV− are the voltagewindow of the positive and negative electrodes, respectively.)FTIR spectrum and SEM images of RGOF are shown inSupporting Information Figure S13a,b. According to test theelectrochemical performance of the RGOF by using a three-electrode system, the gravimetric specific capacitance of theRGOF negative electrode is 205 F g−1 at a current density of 1A g−1 (Supporting Information Figure S14a) based on the eq 3,the mass ratio of negative electrode to positive electrode isabout 5.5:1. Supporting Information Figure S14b compares theCV curves of the negative and positive electrodes at the samescan rate of 10 mV s−1. It clearly shows that the internal CVarea of the negative electrode is almost the same with that ofthe positive electrode, which indicates that the mass ratio issuitable. A series of CV measurements under different potentialwindows were used to determine the best operating voltage ofthe asymmetric supercapacitor (Supporting Information FigureS14c). When the operating window is 1.4 V, the presence ofredox peaks indicate that the pseudocapacitive properties of theasymmetric supercapacitor come from the positive electrode.When the operating voltage was increased to 1.8 V, morefaradic reactions occurred (larger internal CV area). When the

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2041

operating voltage was extended to 2.0 V, there is the emergenceof oxygen.52

To further evaluate the application of the asymmetricsupercapacitor, we carried out the CV and C-DC tests. Unlikethe three-electrode electrochemical feature of CNG@NCHelectrode, the asymmetric supercapacitor displays CV curves ofthe combination of both pseudocapacitive and EDL types(Figure 4a), displaying a capacitive behavior and a desirable fastcharge/discharge property. The voltage is 1.8 V, which isalmost the twice of the conventional active carbon-basedsymmetric capacitors in aqueous solution electrolyte (∼1.0V).48 The galvanostatic discharge curves of the asymmetricsupercapacitor at different current densities were tested inFigure 4b. According to eq 1, the gravimetric capacitance basedon the total mass of active materials (CNG@NCH and RGOF)is 175 F g−1 at a current density of 0.5 A g−1, and the ratecapability of the asymmetric supercapacitor is also given inSupporting Information Figure S15a (131 F g−1 at the currentdensity of 10 A g−1). On the basis of the calculated data,Ragone plot of the asymmetric supercapacitor was obtained andshown in Figure 4c. Data of traditional EDLCs and lithium ionbatteries (LIBs) were provided for comparison and thedensities are based on the weight of active materials.6,48,53

The energy and power densities (E and P) were calculatedaccording to eq 2. The asymmetric supercapacitor exhibits amaximum energy density of 78.75 Wh kg−1 (at a power densityof 473 W kg−1) and a maximum power density of 8424.25 Wkg−1 (at an energy density of 58.51 Wh kg−1). Compared withthe traditional EDLCs, its energy density is much higher at thesame power density level. Moreover, compared with traditionalLIBs, it has a comparable energy density at the same powerdensity level, while it has a much larger power density at the

same energy density level. Compared with other full super-capacitors that have been reported before, the presentasymmetric supercapacitor CNG@NCH//RGOF exhibitssuperior performance (Supporting Information Table S3). Wefurther obtain the volumetric capacitance, volumetric energy,and power densities of our asymmetric supercapacitor atdifferent current densities (Supporting Information FigureS16). The total volume of our asymmetric supercapacitor isbased on the nickel foam and reduced-graphene-oxide film. Inaddition, the asymmetric supercapacitor also exhibits anexcellent cycling stability, which is one of the most importantrequirements in the practical application. After 5000 cycles at acurrent density of 10 A g−1, 99.15% of its capacitance wasretained (Figure 4d). The EIS measurement was used toinvestigate the resistance change of the asymmetric super-capacitor before and after 5000 cycles. From the Nyquist plots,it clearly reveals that the asymmetric supercapacitor exhibits noobvious resistance change after 5000 cycles as compared to theinitial cycle and only a slight increase of Rct from 0.31 to 0.33 Ωwas observed (Supporting Information Figure S15b). The EISresults further demonstrate the exceptional stability of theasymmetric supercapacitor.In summary, through a facile modified hydrothermal method

followed by two-step calcination we synthesized a uniform andunique morphology of cobalt oxide nanobrush, which has notbeen reported before, and then we designed and successfullysynthesized the interwoven 3D frame architecture of CNG@NCH using the optimized electrodeposition method. Thesubsequent electrochemical tests of this interwoven 3D framearchitecture electrode demonstrate that our CNG@NCHelectrode exhibits high comprehensive electrochemical per-formance: high gravimetric and areal specific capacitances, good

Figure 4. (a) CV curves of the asymmetric supercapacitor at different scan rates. (b) Discharge curves of the asymmetric supercapacitor at differentcurrent densities. (c) Ragone plot of the present asymmetric supercapacitor. EDLCs and LIBs are included for comparison. (d) Cycling performanceof the present asymmetric supercapacitor at 10 A g−1 for 5000 cycles. The inset shows the last ten cycles of the asymmetric supercapacitor.

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2042

rate capability, excellent reversibility, high energy density, andsuperior cycling stability, which successfully overcomes thetremendous challenge for pseudocapacitor electrode materials.Moreover, a two-electrode asymmetric supercapacitor based onCNG@NCH was fabricated and it delivers a high specificenergy density and a good rate capability, as well as an excellentcycling stability. Furthermore, the present work meets very wellthe demands of electrode materials with high comprehensiveelectrochemical performance for pseudocapacitors. A moreappropriate negative electrode with high capacitance andsuitable redox reaction is expected to further improve theperformance of this 3D frame architecture electrode materialsbased asymmetric supercapacitor.

■ ASSOCIATED CONTENT*S Supporting InformationAdditional information, experimental section, tables, andfigures. This material is available free of charge via the Internetat http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: (C.H.) [email protected].*E-mail: (L.M.) [email protected].

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was supported by the National Basic ResearchProgram of China (2013CB934103, 2012CB933003), theNational Natural Science Foundation of China (51072153,51272197, 51302203), the International Science & TechnologyCorporation Program of China (2013DFA50840), NationalNatural Science Fund for Distinguished Young Scholars(51425204), Hubei Province Natural Science Fund forDistinguished Young Scholars (2014CFA035), and theFundamental Research Funds for the Central Universities(WUT: 143201003). We thank Professor C. M. Lieber ofHarvard University, Professor Dongyuan Zhao of FudanUniversity, and Dr Jun Liu of Pacific Northwest NationalLaboratory for strong support and stimulating discussions. Weare grateful to Professor J. L. Xie and Dr. X. Q. Liu of theCenter for Materials Research and Analysis of WuhanUniversity of Technology for materials characterization.

■ REFERENCES(1) Miller, J. R.; Simon, P. Science 2008, 321, 651.(2) Conway, B. E. Electrochemical Supercapacitors: ScientificFundamentals and Technological Applications; Kluwer Academic/Plenum: New York, 1999.(3) Winter, M.; Brodd, R. J. Chem. Rev. 2004, 104, 4245.(4) Arico,̀ A. S.; Bruce, P.; Scrosati, B.; Tarascon, J. M.; VanSchalkwijk, W. Nat. Mater. 2005, 4, 366−377.(5) Wang, G.; Zhang, L.; Zhang, J. Chem. Soc. Rev. 2012, 41, 797.(6) Simon, P.; Gogotsi, Y. Nat. Mater. 2008, 7, 845.(7) Zhao, Y.; Liu, J.; Hu, Y.; Cheng, H.; Hu, C.; Jiang, C.; Jiang, L.;Cao, A.; Qu, L. Adv. Mater. 2013, 25, 591.(8) Wu, H.; Geng, J.; Wang, Y.; Wang, Y.; Peng, Z.; Zheng, G. F.Nanoscale 2014, 6, 15316.(9) Simon, P.; Gogotsi, Y. Acc. Chem. Res. 2012, 46, 1094.(10) Zhu, Y.; Murali, S.; Stoller, M. D.; Ganesh, K. J.; Cai, W.;Ferreira, P. J.; Pirkle, A.; Wallace, R. M.; Cychosz, K. A.; Thommes,M.; Su, D.; Stach, E. A.; Ruoff, R. S. Science 2011, 332, 1537.

(11) Wang, Y.; Zhou, T.; Jiang, K.; Da, P.; Peng, Z.; Tang, J.; Kong,B.; Cai, W.; Yang, Z.; Zheng, G. F. Adv. Energy Mater. 2014,DOI: 10.1002/aenm.201470082.(12) Wang, G.; Liu, H.; Horvat, J.; Wang, B.; Qiao, S.; Park, J.; Ahn,H. Chem.Eur. J. 2010, 16, 11020.(13) Zhi, M.; Xiang, C.; Li, J.; Li, M.; Wu, N. Nanoscale 2013, 5, 72.(14) Chen, S.; Duan, J.; Jaroniec, M.; Qiao, S. J. Mater. Chem. A 2013,1, 9409.(15) Meng, Y.; Zhao, Y.; Hu, C.; Cheng, H.; Hu, Y.; Zhang, Z.; Shi,G.; Qu, L. Adv. Mater. 2013, 25, 2326.(16) Li, H.; Yu, M.; Wang, F.; Liu, P.; Liang, Y.; Xiao, J.; Wang, C.;Tong, Y.; Yang, G. Nat. Commun. 2013, 4, 1894.(17) Hu, C.; Chen, J.; Chang, K. J. Power Sources. 2013, 221, 128.(18) Choi, B.; Yang, M.; Jung, S.; Lee, K.; Kim, J.; Park, H. S.; Park,T. J.; Lee, S. B.; Han, Y.; Huh, Y. S. ACS Nano 2013, 7, 2453.(19) Wang, H.; Casalongue, H. S.; Liang, Y.; Dai, H. J. Am. Chem. Soc.2010, 132, 7472.(20) Huang, L.; Chen, D.; Ding, Y.; Feng, S.; Wang, Z.; Liu, M. NanoLett. 2013, 13, 3135.(21) Li, J.; Yang, M.; Wei, J.; Zhou, Z. Nanoscale 2012, 4, 4498.(22) Yuan, C.; Zhang, X.; Hou, L.; Shen, L.; Li, D.; Zhang, F.; Fan,C.; Li, J. J. Mater. Chem. 2010, 20, 10809.(23) Zhou, W.; Xu, M.; Zhao, D.; Li, H. Microporous MesoporousMater. 2009, 117, 55.(24) Guan, C.; Liu, J.; Cheng, C.; Li, H.; Li, X.; Zhou, W.; Zhang, H.;Fan, H. Energy Environ. Sci. 2011, 4, 4496.(25) Shang, C.; Dong, S.; Wang, S.; Xiao, D.; Han, P.; Wang, X.; Gu,L.; Cui, G. L. ACS Nano 2013, 7, 5430.(26) Jiang, X. C.; Tian, B. Z.; Xiang, J.; Qian, F.; Zheng, G. F.; Wang,H. T.; Mai, L. Q.; Lieber, C. M. Proc. Natl. Acad. Sci. U.S.A. 2011, 108,12212.(27) Kim, S. K.; Day, R. W.; Cahoon, J. F.; Kempa, T. J.; Song, K. D.;Park, H. G.; Lieber, C. M. Nano Lett. 2012, 12, 4971.(28) Zhu, G. Y.; He, Z.; Chen, J.; Zhao, J.; Feng, X. M.; Ma, Y. W.;Fan, Q. L.; Wang, L. H.; Huang, W. Nanoscale 2014, 6, 1079.(29) Chang, L.; Pell, W.; Conway, B. E. Electrochim. Acta 1997, 42,3541.(30) Zhang, F.; Yuan, C. Z.; Zhu, J.; Wang, J.; Zhang, X.; Lou, X. W.Adv. Funct. Mater. 2013, 23 (31), 3909.(31) Meher, S. K.; Justin, P.; Rao, G. R. Nanoscale 2011, 3, 683.(32) Niu, Z. Q.; Luan, P. S.; Shao, Q.; Dong, H. B.; Li, J.; Chen, J.;Zhao, D.; Cai, L.; Zhou, W.; Chen, X.; Xie, S. Energy Environ. Sci.2012, 5, 8726.(33) Yang, P. H.; Ding, Y.; Lin, Z. Y.; Chen, Z.; Li, Y.; Qiang, P.;Ebrahimi, M.; Mai, W. J.; Wong, C.; Wang, Z. L. Nano Lett. 2014, 14,731.(34) Hu, L. B.; Chen, W.; Xie, X.; Liu, N.; Yang, Y.; Wu, H.; Yao, Y.;Mauro, P.; Husam, N. A.; Cui, Y. ACS Nano 2011, 5, 8904−8913.(35) Xu, X.; Peng, X.; Jin, H. Y.; Li, T. Q.; Zhang, C.; Gao, B.; Hu, B.;Huo, K. F.; Zhou, J. Adv. Mater. 2013, 25, 5091.(36) Yu, G.; Xie, X.; Pan, L.; Bao, Z.; Cui, Y. Nano Energy 2013, 2,213.(37) Hercule, K. M.; Wei, Q.; Khan, A. M.; Zhao, Y.; Tian, X.; Mai, L.Nano Lett. 2013, 13, 5685.(38) Peng, Z.; Jia, D. S.; Tang, J.; Wang, Y. C.; Wang, Y. H.; Zhang,L. J.; Zheng, G. F. J. Mater. Chem. A 2014, 2, 10904−10909.(39) Preparation of the interwoven 3D frame architecture of CNG@NCH: First, 0.291 g of Co(NO3)2·6H2O and 0.3 g of urea weredissolved in 70 mL of deionized water to form homogeneous pinksolution. The solution was then transferred into a 100 mL Teflon-linedstainless steel autoclave lines with a piece of clean nickel foam (20 mm× 60 mm × 0.4 mm) immersed into the reaction solution. Theautoclave was sealed and maintained at 120 °C for 9 h and then cooleddown to room temperature. The as-synthesized samples were takenout, ultrasonically cleaned at 40 Hz for 5 min in the distilled water andrinsed with ethanol for several times, dried at 70 °C for 8 h, andannealed in Ar gas for 5 h at 450 °C and then in the air for 5 h at 300°C. Second, the as-obtained Co3O4 NB grown on the nickel foam (1cm × 1 cm) works as the working electrode in aqueous electrolyte

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2043

solution of graphene (the graphene was synthesized through amodified Hummer method) and 0.1 M Na2HPO4 in a potential rangeof −2∼0 V at a constant current of 1 mA at 25 °C to synthesize Co3O4NB-graphene (the saturated calomel electrode works as the referenceelectrode, Pt wire works as the counter electrode). Lastly, the as-obtained Co3O4 NB-graphene arrays grown on the nickel foam wasused as the working electrode, saturated calomel electrode works asthe reference electrode, and Pt wire is the counter electrode. Theelectrolyte is 0.05 M Co(NO3)2·6H2O and 0.025 M Ni(NO3)2·6H2O.The CoxNi2x(OH)6x nanoflakes were deposited at a current of 5 mAfor 4 min. The mass of the electrode is 2.0 mg cm−2. As controlexperiments, the hybrid electrode (which the NCH is coated on thesurface of CNG) and CN@NCH were synthesized by changing theconcentration of the electrolyte into 0.02 M Co(NO3)2·6H2O and0.01 M Ni(NO3)2·6H2O and without the electrodeposition ofgraphene, while other conditions were the same as the syntheticprocess of 3D CNG@NCH electrode. Synthesis of reduced-graphene-oxide-film (RGOF): Graphene oxide dispersion (20 mL) made fromthe Hummer method was mixed with 80 mL of distilled water, 0.08mL of hydrazine (85% in water), and 0.35 mL of ammonia (28% inwater) solutions were mixed in a glass beaker. After being vigorouslyshaken for a few minutes, the beaker was put in a water bath (90 °C)for 3 h to get the chemically converted graphene (CCG). Then, 25 mLas-obtained CCG dispersion was vacuum filtrated through a mixedcellulose eater filter membrane (0.22 μm pore size). The vacuum wasdisconnected immediately once no free CCG dispersion was left onthe filtrate cake. The CCG hydrogel films were then carefully peeledoff from the filter membrane and then immediately transferred to aPetri dish and immersed in water overnight to further remove theremaining ammonia and hydrazine.(40) Hummers, W. S., Jr; Offeman, R. E. J. Am. Chem. Soc. 1958, 80,1339.(41) Liang, Y.; Li, Y. G.; Wang, H. L.; Zhou, J. G.; Wang, J.; Regier,T.; Dai, H. J. Nat. Mater. 2011, 10, 780.(42) Si, Y. C.; Samulski, E. T. Nano Letter. 2008, 8, 1681.(43) Guo, H. L.; Wang, X. F.; Qian, Q. Y.; Wang, F. B.; Xia, X. H.ACS Nano 2009, 3, 2653.(44) Kim, G.; Nam, I.; Kim, N. D.; Park, J. S.; Park, S. M.; Yi, J.Electro. Commun. 2012, 22, 93.(45) Mai, L.; Yang, F.; Zhao, Y.; Xu, X.; Xu, L.; Luo, Y. Nat. Commun.2011, 2, 381.(46) Characterization: The XRD measurements were performed byusing a Bruker D8 Advance X-ray diffractometer with a non-monochromated Co Ka X-ray source. The Fourier transform infraredspectra were recorded by using 60-SXB infrared spectrometer. SEMimages were collected by using a JEOL JSM-7100F at an accelerationvoltage of 20 kV. TEM and HRTEM images were recorded by using aJEOL JEM-2100F STEM/EDS microscope. The Oxford EDS IE250 isused for recording the EDS. Atomic Absorption Spectra wasperformed by using a GBC AVANTA M Atomic absorptionspectrometer. Electrochemical measurement: In the three-electrodesystem, the nickel foam supported active materials as the workingelectrode (typically 1 cm × 1 cm) and 1 M LiOH aqueous solution asthe electrolyte with Pt wire and Ag/AgCl (saturated KCl) electrodes ascounter and reference electrodes, respectively. The distance betweenthe working electrode and the counter electrode was about 2 cm. Theasymmetric supercapacitor was measured with a two-electrode systemwith the 3D frame architecture CNG@NCH as the positive electrode,the RGOF as the negative electrode, and 1 M LiOH aqueous solutionas the electrolyte. All the electrochemical measurements were carriedout with an autolab electrochemical workstation.(47) Mai, L.; Khan, A. M.; Tian, X.; Hercule, K. M.; Zhao, Y.; Xu, L.;Xu, X. Nat. Commun. 2013, 4, 2923.(48) Zhou, C.; Zhang, Y.; Li, Y.; Liu, J. P. Nano Lett. 2013, 13, 2078.(49) Hou, Y.; Cheng, Y.; Hobson, T.; Liu, J. Nano Lett. 2010, 10,2727.(50) Yang, X.; Cheng, C.; Wang, Y.; Qiu, L.; Li, D. Science 2013, 341,534.(51) Yang, X.; Zhu, J.; Qiu, L.; Li, D. Adv. Mater. 2011, 23, 2837.

(52) Ji, J. Y.; Zhang, L.; Ji, H. X.; Li, Y.; Zhao, X.; Bai, X.; Fan, X. B.;Zhang, F. B.; Ruoff, R. S. ACS Nano 2013, 7, 6242.(53) El-Kady, M. F.; Strong, V.; Dubin, S.; Kaner, R. B. Science 2012,335, 1326.

Nano Letters Letter

DOI: 10.1021/nl504901pNano Lett. 2015, 15, 2037−2044

2044